684 



Fishery Bulletin 97(3), 1999 



Nevada) and 50^N, 170°W (Aleutian Islands) (NOAA^). 

 A plot of the data revealed an increase in the inten- 

 sity of the pressure differences during the winter 

 months. Therefore, annual means were calculated 

 from the monthly means for November to March. 



4.2 



3.8 



3.4 1 



3.0  



2.6 



2.2 



2.2 



Female 

 Cordell Bank 



3.8  



3.4 



C7> 



i 3.0 

 o 



o 



2.6 



22 



3,8 



3.4 



3.0 - 



2.6 



Female 

 Eureka 



Male 

 Eureka 



— Combined 

 Age1 



-— Age 2 



- - Age 3 



82 



83 

 Year 



84 



Figure 4 



Comparison of mean annual otolith growth indices 11982-84) 

 among models with and without interaction terms for yellow- 

 tail rockfish. Bars represent ±1 standard error of the mean. 



Owing to strong interdependence among these 13 

 oceanographic variables, a principal component 

 analysis (SAS Institute Inc., 1988) was conducted to 

 reduce the dimensionality of the information and to 

 insure statistical independence of the data. The or- 

 dination was conducted on the correlation ma- 

 trix, so that each variable was weighted equally. 

 The otolith data were then compared with the prin- 

 cipal components of the oceanographic data to quan- 

 tify the impact of the 1983 El Nifio event on the 

 growth of yellowtail and widow rockfish otoliths. 



Results 



Patterns of annual otolith growth were similar 

 in widow and yellowtail rockfish from 1980 to 

 1987 (Fig. 5). A 12.67r and 20.5% reduction in 

 otolith growth occurred during 1983 for widow 

 and yellowtail rockfish, respectively. This year 

 of reduced otolith growth was followed by en- 

 hanced growth, especially for yellowtail rockfish. 

 The growth estimates calculated for 1980 and 

 1987 were based on samples from one year, i.e. 

 the 1979 and 1983 year classes respectively, and 

 these estimates should be viewed with caution. 

 Although most of the variance in otolith growth 

 is explained by the year effect, the other effects 

 in the model are worth noting (Table 2). There 

 was a significant difference in otolith growth be- 

 tween sexes for yellowtail rockfish (i.e. otoliths 

 from females grew faster than those from males); 

 however, this was not the case for widow rock- 

 fish. Growth of the otolith at age differs between 

 species (Fig. 6), which may be the result of so- 

 matic growth rate differences between the spe- 

 cies. In a simple simulation, similar curves were 

 produced by varying the von Bertalanffy growth 

 coefficient (k). Yellowtail rockfish exhibit faster 

 growth than do widow rockfish, thus yielding 

 relatively lower indices of otolith growth with in- 

 creasing age. There was a significant difference 

 in otolith growth between the two ports for both 

 species (i.e. otoliths from fish collected in Eureka 

 grew faster than those from Bodega). Further 

 analysis revealed that these differences began in 

 1983. Figure 7 shows results of four separate 

 ANOVAs run on each of the four species-port com- 

 binations. In these analyses, sex and port were 

 excluded as effects in the model. Otolith growth 

 was similar at the two ports prior to 1983. AI- 



^ Pacific Marine Environmental Laboratory, National Oceanic 

 and Atmospheric Administration, Seattle, Washington. 

 1994. Unpubl. data. 



